- Email: [email protected]
Gas chromatography of the coke formed in catalytic cracking reactions
360 CHROM.
NOTES
5722
Gas chromatqjraphy
of *he coke formed
in catalytic
crackiog reactions
There have been several research reports regarding the formation of coke, the amount of coke on the catalyst and the relative coking ability of the various hydrocarbons; THOMAS*managed to extract 27% of the’deposit formed by the passage of gz-octenes over an alumina-zirconia-silica catalyst at 375” ‘and at about 8 atm with benzene. Part of the extract (40%) was soluble in acetone and was a dark oil. The remaining portion of the extract, insoluble in acetone, was a dark solid melting at rgg” to 210~. Hydrogen to carbon ratios were different for the two portions of the extract. Under standardized cracking conditions different amounts of coke are formed from various hydrocarbons. The relative amounts of coke formed from various hydrocarbons under constant conditions are given by VOGE et ~11.2,Using a radioactive tracer technique MCMAHON~ concluded that all carbon atoms take part, with equal facility,in coke formation when cracking n-heptane. The same thing was also observed with noctane and a-octene. MELIK-ZADEet a? .a-@ observed the:formation &coke by passing 14C-labelled rt-pentylbenzene over a catalyst both alone and in w-hexadecane. Their observations indicated that most of the coke was formed from the reaction with the npentyl group and the contribution of the benzene-ring to the formation of coke was very low. HIGHTOWERAND EMMETT’ by using radioactive tracers have concluded that the olefins are the most active compounds in forming coke. Alkyl aromatics, benzene and paraffins follow olefins in the order.given, in their ability to form coke. There are practically no reports in the literature describing the analysis of coke. If the various hydrocarbons differ in their ability to form coke it seems probable that the coke formed from them will have a different composition and, that this composition may also change even for the satie hydrocarbon if the cracking conditions vary. With this in mind, an analytical study of coke formed in the catalytic cracking of n-pentane and isopentane was carried out and the results obtained on the gas chromatography of the coke are now presented. Exjbevimental The experimental set-up is similar to the usual apparatus for catalytic hydrocarbon (vapour) cracking and consists of a helium carrier gas supply, a hydrocarbon bubble saturator, a Pyrex glass reactor mounted vertically containing 25 g of Davisong8o S-Al catalyst (7~x0 mesh). The reactor is placed in a furnace in which a temperature of 518 =t: IO is maintained by a temperature controller. The apparatus also contains a catalyst regeneration system in which the catalyst is regenerated by passing water vapour and air through it at 520~. A stream of helium at a pressure of one atmosphere saturated with hydrocarbon vapour passes at a rate of 40 ml/min through the catalyst. Thegas emerging from the reactor could be injected into the gas chromatograph, by means of a by-pass system, and analysed. After the hydrocarbon vapour had been passed over the catalyst for I h, the saturator was by-passed and the helium flow continued through the reactor to remove any adsorbed, unabsorbed and uncoked by-products. At the same time the heating current was turned off. ,After the reactor hid,,cooled down it was’ disconnected from the system and the catalyst with its coke deposit was taken out and was placed in pure acetone in a bottle overnight. After J. CA~omalogv.,’ 64
(1972)
360-363
NOTES
Pig. I'.Analysis of (a) coke formed in w-pontsne cracking ing. Column : 6 ft. x & in., 20% SE-30 on Chromosorb; tcctor.
361
and (b) colcc found in isopcntsnc cbA~tompcrature, 2Go"; flame ionization de-
filtering this, a clear brownish yellow solution of coke dissolved in acetone was obtained. (Joke ‘solutions obtained in this manner from several’ runs were collected together. A concentrated solution of ,coke in acetone wasinjected into a gas chroma‘, togr$ph fat analysis, Fig; I; shows ‘a ‘chromatogram obtained from a coke ‘solution, the coke being obtained by’ cracl&ig n-pent&e. Pig, rb’is a sim’ilar ‘chromatogram obtaineli with isopentane coke. Obviously the two chromatograms show some ‘differences.” A first concltision from this analysis is t&t;’ coke obtained from’ different compoufids ‘&ff ers in :c&ii$osition. The conditions of cracking yere’ controlled during ‘all ru&. It was assumed, therefore, that coke obtained in iririo’us runs ‘ior ‘the’ same hydrocarbon cracking bad a constant composition; The ,yields,of,coke from different runs were collected together since a very ,small amount of.coke is obtained in &individual J.
Ch~otrzologv., 64 (1972) 360-364
362
NOTES
run it cannot be analysed separately by gas chromatography. The two chromatograms represent only that portion of the coke soluble in acetone. A rough measure of what, percentage of coke is soluble in acetone was obtained by a series of experiments with rc-pentane as the specimen hydrocarbon. After dissolving the coke in acetone, the dried catalyst was replaced in the reactor and the helium flow started. The temperature of the reactor furnace was grgdually increased in ‘order that all acetone was removed from the catalyst. Any acetone remaining on the heated catalyst might produce its own effects on the catalyst. If radioactive acetone tracer was used it would be easy to see if any acetone still remained on the catalyst after passing helium and gradually increasing the temperature to the desired value. In the present experiments it was assumed that all the acetone is removed by passing helium for I h
18
2
16
21
g 14.II P 125
g’” $ 8b 6-
,
I
I
I
ABC
ABC
State
.I
A&C
ot the catalyst
Fig. 2, Effect of dissolying coke from catalyst by acetone on the activity of catalyst. 0, Rcgcnoratcd catalyqt; 8, c@alygt from which coke is dissolved: 0, lowest activity when coke is completely deposited dn catalyst. The upper two points are after 4 min from start. State of the catalyst: A =t regenerated catalyst, B = catalyst from which coke is dissolved, and C = catalyst corn: plctely covorcd with coke.
and gradual19 increasing the temperature. The catalyst was maintained ,at 514 & 1’ and the activity,of it for a-pentane cracking measured at 4 min and 12’qin after tipentane vapbur started, and continued until constant activity was ,obtained. In ‘the next run the catalyst was regenerated by passage of water vapdur and’air’ over the catalystat si&for one day and the,activity of this regenerated catalyst was measured at the ‘same:’intervals: ,of time. These two experiments were ‘repeFted several times. Activity is, given- in >terms of per cent conversion of +z-pentane.’ Per cent conversion w,as’ calculate$,~by, determining ,the, amount ‘of n-pen&me before ,and after cracking; This ,wasdone by .measuring th:e height of Iz-pentarie peaks before ,and Tfter cracking. :,,“I ‘. :
,,“’
,,”
;,
.’
.”
i.
.y:
,”
,, :,,:.Per cent conversion =
‘C&&.l*gY.,
’
‘.‘,
t&
(ig72)
360-363
decrease in partial. pressure. of. +z-pentane ‘x Ioo original partial pressure of +pentane t ” ’
.,’
NOTES
363
Fig. 2 represents the data Every curve SLOWS three points.
from several experiments under these two conditions. The empty circles represent the values obtained with regenerated catalyst, half solid ‘circles indicate the values with catalyst fr.om which coke has been removed by acetone and the solid circles represent the constant lov.&t value of per cent conversion obtainable at the end of every run. During the regeneration process activity is raised’from the value at the lowest point to the value at the highest point. If we assume that by this regeneration process recovery of activity is IOO~/~ (i.e. all coke is removed from the catalyst), the recovery of activity by the use of acetone (corresponding to middle points) could be calculated from the per cent conversion values at the three points. It was found that an increase in activity of approximately 16% takes place after using acetone. A rough estimate is therefore that only ~(3% of coke is dissolved in the acetone. Further work on naming the compounds in coke and the study of coke obtained from olefins, aromatics, etc. is proposed. The thin layer chromatography of these cokes is being carried out. The author wishes to thank Dr. A. SOLRAKKEN for his constant encouragement and
valuable
suggestions
throughout
this work.
Thanks
are also due to Professor
P. II. EMMETT, The John Hopkins University, U.S.A. for showing interest in this work and. a number of valuable suggestions given by him during his visit ,to this institute. The authoiis grateful to the NORAD, Oslo, for the flnaricial support during this work. K. H. KARMARKAR*
~m~itute of lndi&&l Cht.m+stvy, Techngcal Jlutivevsity of Norway, Trondheim ‘(Norway) 'C.L. TWOM&,
,I.'Amey. Ckcm.
Sm.,
GG‘ (1944) 1586.
I-I. M. VOGE; G. M. GOOD AND EL S. GRIZENSBELDIZR, Pyoc. Woyld Petyolcztm Coqwss, gvd, Tho Hagwe, rgp, Sect. IV, p. 124. R. E, MCMAHON, Ind. Bng. Cl&em., 47 (1955) 844, M. M, MELIK-ZADE, M. R. MUSAW AND I. G. SAFARALIEVA, Azcyb. Nejt, ICJtoz., 39, No. 8 (19GO) 33;
C.A.,
55 (I9GI)
25225.
M. M. MELIIC-ZADE, M. R. MUSAEV AND I. G. SAFARALIEVA,
AZ+.
Neft.
KJtoz.,
39, No. II
., ‘55
M.
MELIX-ZADE AND M. R. MTJSAEV, Azevb, Ncfl. KJzoz., 40, No. 2 (I~GI) 42 : CA (, 56 (1~62)
fP*%.
I-IIGHTOWE,R AND P. I-1, EMMETT, J, Amer.
i
Chew.
SOL, 87 (1,965)
939.
First received July sth, 1971; revised manuscript received September zIst, rg7r l
Pcrmancnt
nddrcss: Chemistry
Dopczrtmont, PoonsL University,
Poonw7,
Indict.
J, CJtvbinalogr.~, 64’ (1672) 360-363
NOTES
5722
Gas chromatqjraphy
of *he coke formed
in catalytic
crackiog reactions
There have been several research reports regarding the formation of coke, the amount of coke on the catalyst and the relative coking ability of the various hydrocarbons; THOMAS*managed to extract 27% of the’deposit formed by the passage of gz-octenes over an alumina-zirconia-silica catalyst at 375” ‘and at about 8 atm with benzene. Part of the extract (40%) was soluble in acetone and was a dark oil. The remaining portion of the extract, insoluble in acetone, was a dark solid melting at rgg” to 210~. Hydrogen to carbon ratios were different for the two portions of the extract. Under standardized cracking conditions different amounts of coke are formed from various hydrocarbons. The relative amounts of coke formed from various hydrocarbons under constant conditions are given by VOGE et ~11.2,Using a radioactive tracer technique MCMAHON~ concluded that all carbon atoms take part, with equal facility,in coke formation when cracking n-heptane. The same thing was also observed with noctane and a-octene. MELIK-ZADEet a? .a-@ observed the:formation &coke by passing 14C-labelled rt-pentylbenzene over a catalyst both alone and in w-hexadecane. Their observations indicated that most of the coke was formed from the reaction with the npentyl group and the contribution of the benzene-ring to the formation of coke was very low. HIGHTOWERAND EMMETT’ by using radioactive tracers have concluded that the olefins are the most active compounds in forming coke. Alkyl aromatics, benzene and paraffins follow olefins in the order.given, in their ability to form coke. There are practically no reports in the literature describing the analysis of coke. If the various hydrocarbons differ in their ability to form coke it seems probable that the coke formed from them will have a different composition and, that this composition may also change even for the satie hydrocarbon if the cracking conditions vary. With this in mind, an analytical study of coke formed in the catalytic cracking of n-pentane and isopentane was carried out and the results obtained on the gas chromatography of the coke are now presented. Exjbevimental The experimental set-up is similar to the usual apparatus for catalytic hydrocarbon (vapour) cracking and consists of a helium carrier gas supply, a hydrocarbon bubble saturator, a Pyrex glass reactor mounted vertically containing 25 g of Davisong8o S-Al catalyst (7~x0 mesh). The reactor is placed in a furnace in which a temperature of 518 =t: IO is maintained by a temperature controller. The apparatus also contains a catalyst regeneration system in which the catalyst is regenerated by passing water vapour and air through it at 520~. A stream of helium at a pressure of one atmosphere saturated with hydrocarbon vapour passes at a rate of 40 ml/min through the catalyst. Thegas emerging from the reactor could be injected into the gas chromatograph, by means of a by-pass system, and analysed. After the hydrocarbon vapour had been passed over the catalyst for I h, the saturator was by-passed and the helium flow continued through the reactor to remove any adsorbed, unabsorbed and uncoked by-products. At the same time the heating current was turned off. ,After the reactor hid,,cooled down it was’ disconnected from the system and the catalyst with its coke deposit was taken out and was placed in pure acetone in a bottle overnight. After J. CA~omalogv.,’ 64
(1972)
360-363
NOTES
Pig. I'.Analysis of (a) coke formed in w-pontsne cracking ing. Column : 6 ft. x & in., 20% SE-30 on Chromosorb; tcctor.
361
and (b) colcc found in isopcntsnc cbA~tompcrature, 2Go"; flame ionization de-
filtering this, a clear brownish yellow solution of coke dissolved in acetone was obtained. (Joke ‘solutions obtained in this manner from several’ runs were collected together. A concentrated solution of ,coke in acetone wasinjected into a gas chroma‘, togr$ph fat analysis, Fig; I; shows ‘a ‘chromatogram obtained from a coke ‘solution, the coke being obtained by’ cracl&ig n-pent&e. Pig, rb’is a sim’ilar ‘chromatogram obtaineli with isopentane coke. Obviously the two chromatograms show some ‘differences.” A first concltision from this analysis is t&t;’ coke obtained from’ different compoufids ‘&ff ers in :c&ii$osition. The conditions of cracking yere’ controlled during ‘all ru&. It was assumed, therefore, that coke obtained in iririo’us runs ‘ior ‘the’ same hydrocarbon cracking bad a constant composition; The ,yields,of,coke from different runs were collected together since a very ,small amount of.coke is obtained in &individual J.
Ch~otrzologv., 64 (1972) 360-364
362
NOTES
run it cannot be analysed separately by gas chromatography. The two chromatograms represent only that portion of the coke soluble in acetone. A rough measure of what, percentage of coke is soluble in acetone was obtained by a series of experiments with rc-pentane as the specimen hydrocarbon. After dissolving the coke in acetone, the dried catalyst was replaced in the reactor and the helium flow started. The temperature of the reactor furnace was grgdually increased in ‘order that all acetone was removed from the catalyst. Any acetone remaining on the heated catalyst might produce its own effects on the catalyst. If radioactive acetone tracer was used it would be easy to see if any acetone still remained on the catalyst after passing helium and gradually increasing the temperature to the desired value. In the present experiments it was assumed that all the acetone is removed by passing helium for I h
18
2
16
21
g 14.II P 125
g’” $ 8b 6-
,
I
I
I
ABC
ABC
State
.I
A&C
ot the catalyst
Fig. 2, Effect of dissolying coke from catalyst by acetone on the activity of catalyst. 0, Rcgcnoratcd catalyqt; 8, c@alygt from which coke is dissolved: 0, lowest activity when coke is completely deposited dn catalyst. The upper two points are after 4 min from start. State of the catalyst: A =t regenerated catalyst, B = catalyst from which coke is dissolved, and C = catalyst corn: plctely covorcd with coke.
and gradual19 increasing the temperature. The catalyst was maintained ,at 514 & 1’ and the activity,of it for a-pentane cracking measured at 4 min and 12’qin after tipentane vapbur started, and continued until constant activity was ,obtained. In ‘the next run the catalyst was regenerated by passage of water vapdur and’air’ over the catalystat si&for one day and the,activity of this regenerated catalyst was measured at the ‘same:’intervals: ,of time. These two experiments were ‘repeFted several times. Activity is, given- in >terms of per cent conversion of +z-pentane.’ Per cent conversion w,as’ calculate$,~by, determining ,the, amount ‘of n-pen&me before ,and after cracking; This ,wasdone by .measuring th:e height of Iz-pentarie peaks before ,and Tfter cracking. :,,“I ‘. :
,,“’
,,”
;,
.’
.”
i.
.y:
,”
,, :,,:.Per cent conversion =
‘C&&.l*gY.,
’
‘.‘,
t&
(ig72)
360-363
decrease in partial. pressure. of. +z-pentane ‘x Ioo original partial pressure of +pentane t ” ’
.,’
NOTES
363
Fig. 2 represents the data Every curve SLOWS three points.
from several experiments under these two conditions. The empty circles represent the values obtained with regenerated catalyst, half solid ‘circles indicate the values with catalyst fr.om which coke has been removed by acetone and the solid circles represent the constant lov.&t value of per cent conversion obtainable at the end of every run. During the regeneration process activity is raised’from the value at the lowest point to the value at the highest point. If we assume that by this regeneration process recovery of activity is IOO~/~ (i.e. all coke is removed from the catalyst), the recovery of activity by the use of acetone (corresponding to middle points) could be calculated from the per cent conversion values at the three points. It was found that an increase in activity of approximately 16% takes place after using acetone. A rough estimate is therefore that only ~(3% of coke is dissolved in the acetone. Further work on naming the compounds in coke and the study of coke obtained from olefins, aromatics, etc. is proposed. The thin layer chromatography of these cokes is being carried out. The author wishes to thank Dr. A. SOLRAKKEN for his constant encouragement and
valuable
suggestions
throughout
this work.
Thanks
are also due to Professor
P. II. EMMETT, The John Hopkins University, U.S.A. for showing interest in this work and. a number of valuable suggestions given by him during his visit ,to this institute. The authoiis grateful to the NORAD, Oslo, for the flnaricial support during this work. K. H. KARMARKAR*
~m~itute of lndi&&l Cht.m+stvy, Techngcal Jlutivevsity of Norway, Trondheim ‘(Norway) 'C.L. TWOM&,
,I.'Amey. Ckcm.
Sm.,
GG‘ (1944) 1586.
I-I. M. VOGE; G. M. GOOD AND EL S. GRIZENSBELDIZR, Pyoc. Woyld Petyolcztm Coqwss, gvd, Tho Hagwe, rgp, Sect. IV, p. 124. R. E, MCMAHON, Ind. Bng. Cl&em., 47 (1955) 844, M. M, MELIK-ZADE, M. R. MUSAW AND I. G. SAFARALIEVA, Azcyb. Nejt, ICJtoz., 39, No. 8 (19GO) 33;
C.A.,
55 (I9GI)
25225.
M. M. MELIIC-ZADE, M. R. MUSAEV AND I. G. SAFARALIEVA,
AZ+.
Neft.
KJtoz.,
39, No. II
., ‘55
M.
MELIX-ZADE AND M. R. MTJSAEV, Azevb, Ncfl. KJzoz., 40, No. 2 (I~GI) 42 : CA (, 56 (1~62)
fP*%.
I-IIGHTOWE,R AND P. I-1, EMMETT, J, Amer.
i
Chew.
SOL, 87 (1,965)
939.
First received July sth, 1971; revised manuscript received September zIst, rg7r l
Pcrmancnt
nddrcss: Chemistry
Dopczrtmont, PoonsL University,
Poonw7,
Indict.
J, CJtvbinalogr.~, 64’ (1672) 360-363